Is it possible to predict both the potential for selection and epistatic interactions across a biological network? Here we propose to quantitatively address the physiological basis of adaptation through the integration of experimental and computational approaches. Our model system is one in which the central, essential and highly interconnected metabolic pathway of Methylobacterium has been disabled and replaced with a foreign, unrelated pathway. The unique advantage of this engineered system is that this replacement specifically results in a 3-fold reduction in fitness, growth rate and metabolic flux, as well as 2.5-fold lower yield and a 30-fold redistribution of flux within the central metabolic hub. Because this alteration directly causes sub-optimal performance, we hypothesize that this will focus selection upon this subsystem during experimental evolution such that adaptation will largely proceed through mutations in the substituted pathway and/or those that it physiologically interacts with. Furthermore, we suggest that increasingly extended and verified mathematical models of this metabolic subsystem and its connections to the metabolic network will allow us to make testable predictions of the fitness effects of altering the activity of individual system components, as well as epistatic interactions between enzymes. Our preliminary results support both our model's predictions and the assertion that adaptation will strike this central metabolic hub. Our specific aims are to 1.) explore the potential for selection with metabolic models and directly test predictions by modulating expression levels of enzymes, 2.) evolve replicate populations of the ancestral strain and examine phenotypic and genetic changes throughout the course of adaptation and 3.) test the role of epistasis in the adaptive trajectories observed or synthesized. The result of this project will be a novel model system and conceptual framework to apply a comprehensive, systems biology approach to understanding the physiological basis of selection and epistasis in adaptation. It also represents the opportunity to address adaptation occurring after introduction of new genetic material via horizontal gene transfer. We anticipate that placing selection and epistasis into a quantitative framework will have public health impacts ranging from the adaptation of pathogens, the modeling of metabolic diseases, to prognostic predictions of the 'adaptive' fate of a population of cancer cells with mutated oncogenes and tumor suppressors. [unreadable] [unreadable] [unreadable]